A method for predicting rainfall based on marine radar images

Abstract

The application is a rain prediction method based on marine radar images, belonging to the technical field of radar image processing and meteorological service, comprising the following steps: obtaining a radar original data file, performing azimuth extraction on the obtained radar original data file, performing merging interpolation processing and coordinate conversion processing, calling a display function, and obtaining a radar echo image; the radar echo image comprises a radar rain image and a sunny day radar image; the radar rain image and the sunny day radar image in the radar echo image are differentiated to remove central clutter interference; the radar rain image is reduced to a fixed resolution, and a level set method is used for radar echo image segmentation; the motion vector field is calculated by an optical flow method for continuous N segmented radar echo images; and the rain is predicted by combining the radar echo image and the motion vector field, thereby improving the accuracy of rain prediction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar image processing and meteorological service, and particularly relates to a rainfall prediction method based on a marine radar image. BACKGROUND

[0002] China has a vast land area, including various climate regions, and the climate characteristics are prominent, while various meteorological disasters also affect various places in China. The loss caused by the weather cannot be ignored. Rainfall, as one of the climate characteristics in most areas of China, helps water circulation and promotes crop growth, but at the same time, rainfall, as one of the common factors causing meteorological disasters, may cause disasters such as floods, landslides, and mudslides. Therefore, increasing the research on rainfall can play an important role in flood prevention, strengthening building stability, and reducing losses. With the development of technology, weather radars have replaced traditional rain gauges for rainfall monitoring, and there has been great progress in gradually understanding more complex weather. Meteorological radars are more commonly used in rainfall problems, and although they have advantages in accuracy, the high cost and complex equipment limit the application of meteorological radars. Marine radars can also have advantages in rainfall monitoring. Marine radars are very inexpensive, small in size, and easy to install; they have a certain range resolution and angular resolution; they are sensitive to rainfall and can better collect rainfall images, which can significantly reduce costs when applied to rainfall analysis. For near-weather forecasting methods, there are two main ways: numerical model prediction and radar echo extrapolation. Model prediction has lower accuracy in short-term near-weather prediction, but radar echo extrapolation has better results in short-term prediction. Therefore, marine radar data can be used to predict rainfall conditions. SUMMARY

[0003] In view of the above-mentioned technical problems of high cost and complex equipment, and in view of the characteristics of marine radar echo sequences, the present application adopts a rainfall prediction method based on marine radar images, which comprises the following steps: a rainfall prediction method based on marine radar images, comprising the following steps:

[0004] Obtain a radar original data file, perform azimuth extraction on the obtained radar original data file, perform merging interpolation processing and coordinate conversion processing, call a display function, and obtain a radar echo image; the radar echo image includes a radar rainfall image and a sunny radar image;

[0005] Differencing the radar rainfall image and the sunny radar image in the radar echo image removes center clutter interference;

[0006] The radar rainfall image is reduced to a fixed resolution, and a level set method is used for radar echo image segmentation;

[0007] The motion vector field is calculated by optical flow method on the N continuous segmented radar echo images;

[0008] The rain is predicted by combining the radar echo image and the motion vector field.

[0009] Further, the radar original data file is acquired, the acquired radar original data file is subjected to azimuth extraction, is subjected to merging interpolation processing and coordinate conversion processing, a display function is called, and a radar echo image is obtained; the radar echo image includes a radar rain image and a sunny day radar image; the following steps are included:

[0010] S21: In the case of rain, the marine radar is used, rain data is collected by a collection card, and a radar original data file is acquired;

[0011] S22: The collected radar original data file is subjected to azimuth extraction, and the high 2 bits in the sampling data are azimuth reference and azimuth data respectively; the jump between 0 and 1 of the highest bit is used to represent that the antenna has turned through the azimuth reference, and then the azimuth data information of the next highest bit is extracted, so as to realize the formatting of the collected data information;

[0012] S23: The formatted data information is subjected to merging interpolation processing, and the correspondence between the azimuth pulse and the trigger pulse may be deviated under different ranges; if one azimuth pulse corresponds to multiple trigger pulses, the first echo information is retained; if one azimuth pulse has no corresponding trigger pulse, the last echo information is copied to this azimuth pulse, so as to ensure that each azimuth has corresponding echo information;

[0013] S24: The data subjected to the merging interpolation processing is subjected to coordinate conversion, the radar information is in polar coordinates, the polar coordinates are converted into rectangular coordinates, then matrix data is generated, and the matrix data file is stored;

[0014] S25: Then the matrix data file is read in and saved to the corresponding matrix in the memory, the azimuth distance matrix is converted into Cartesian coordinates, and is stored in a new matrix, and then a display function is called to display the radar echo image.

[0015] Further, the radar rain image and the sunny day radar image in the radar echo image are subjected to difference, including the following steps:

[0016] S31: In the case of sunny day, radar data with the same radar collection range and gain as the radar collection range and gain are collected, and are converted into a radar echo image;

[0017] S32: The radar rain image and the sunny day radar image are subjected to difference processing by corresponding subtraction of pixel values, and only rain area information is left.

[0018] Furthermore, the step of reducing the radar rainfall image to a fixed resolution and segmenting the radar echo image using the level set method includes the following steps:

[0019] S41: Reduce radar rainfall images to a fixed resolution;

[0020] S42: Set the initial parameters λ1, λ2, α1 weighting constant and α2, the maximum number of iterations IterNum, and the edge detection matrix g;

[0021] S43: Initialize the pseudo-level set, setting a portion of the pseudo-level set to u(x) > 0.5, and setting the other portion of the pseudo-level set to u(x) < 0.5;

[0022] S44: Design two fitting images: a globally blurred fitting image and a locally blurred fitting image, and construct an energy function that includes global and fitting terms;

[0023] S45: Calculate the energy of the entire image domain, update the parameters of the entire image domain, calculate the new energy of the entire image domain, change the membership degree, and use a Gaussian filter to regularize and smooth the pseudo-level set function.

[0024] S46: Repeat step S45 until the iteration is completed after IterNum iterations;

[0025] S47: After image segmentation, information about the rainfall area is preserved.

[0026] Furthermore, the calculation of the motion vector field from N consecutive segmented radar echo images using the optical flow method includes the following steps:

[0027] S51: Calculate the optical flow field from N radar echo images using the LK optical flow method;

[0028] S52: Track a sequence of N images to generate a smooth motion vector field.

[0029] Furthermore, N ≥ 10.

[0030] Furthermore, the method of combining radar echo images with motion vector fields to predict rainfall includes the following steps:

[0031] By combining the deterministic components and perturbation terms of Lagrange persistence, precipitation is dynamically scaled.

[0032] An autoregressive (AR) process is used to calculate the evolution of the precipitation field, thereby enabling rainfall prediction.

[0033] A rainfall prediction method based on marine radar imagery includes:

[0034] Acquisition module: Used to acquire raw radar data files;

[0035] The acquisition module is used to extract the azimuth of the acquired raw radar data file, perform merging, interpolation, and coordinate transformation, and call the display function to obtain the radar echo image; the radar echo image includes radar images during rainfall and radar images during clear weather.

[0036] Differential module: used to differentiate the radar rainfall image from the radar echo image with the clear sky radar image to remove central clutter interference;

[0037] Segmentation module: Used to reduce radar rainfall images to a fixed resolution and segment radar echo images using the level set method;

[0038] Calculation module: used to calculate the optical flow field from N consecutive segmented radar echo images using the optical flow method;

[0039] Prediction module: Used to combine radar echo images with motion vector fields to predict rainfall.

[0040] This invention provides a rainfall prediction method based on marine radar images. The method first uses the differential method to remove large clutter interference from the data, then uses the level set method to extract the outline of the rainfall area, and then adds a module to retain the image data to preserve the rainfall area information, which is of great significance for subsequent motion vector field extrapolation. Then, the motion vector field of 10 consecutive radar echo images is calculated using the LK optical flow method. Combining the information from both the radar image and the motion vector field, the rainfall situation 30 minutes later is predicted.

[0041] This invention applies image segmentation methods to rainfall analysis using marine radar, enabling simple rainfall prediction with cost-effective marine radar and reducing the cost of rainfall forecasting. The level set method in this invention solves the problem of extracting useful information after image segmentation and satisfies the conditions of the LK optical flow method when calculating the motion vector field. It can be introduced into the prediction module, improving the accuracy of rainfall prediction. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is the overall flowchart of this method;

[0044] Figure 2 It is a radar echo image of rainfall;

[0045] Figure 3 This is a differential radar echo diagram for both sunny and rainy weather.

[0046] Figure 4 It is a segmentation diagram of radar echo image;

[0047] Figure 5 It is an extrapolation diagram of the motion vector field;

[0048] Figure 6 This is a rainfall forecast map. Detailed Implementation

[0049] It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0052] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of the invention. It should also be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale. Techniques, methods, and devices known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and devices should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters in the following figures denote similar items; therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0053] In the description of this invention, it should be understood that the orientation or positional relationship indicated by directional terms such as "front, back, up, down, left, right", "horizontal, vertical, horizontal" and "top, bottom" is generally based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing this invention and simplifying the description. Unless otherwise stated, these directional terms do not indicate or imply that the device or element referred to must have a specific orientation or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on the scope of protection of this invention. The directional terms "inner" and "outer" refer to the inner and outer contours relative to the outline of each component itself.

[0054] For ease of description, spatial relative terms such as "above," "over," "on the upper surface of," "above," etc., are used herein to describe the spatial positional relationship of a device or feature as shown in the figures to other devices or features. It should be understood that spatial relative terms are intended to encompass different orientations in use or operation besides the orientation of the device as described in the figures. For example, if the device in the figures is inverted, a device described as "above" or "above" other devices or structures would subsequently be positioned as "below" or "under" other devices or structures. Thus, the exemplary term "above" can include both "above" and "below." The device may also be positioned in other different ways (rotated 90 degrees or in other orientations), and the spatial relative descriptions used herein will be interpreted accordingly.

[0055] Furthermore, it should be noted that the use of terms such as "first" and "second" to define components is merely for the purpose of distinguishing the corresponding components. Unless otherwise stated, the above terms have no special meaning and therefore should not be construed as limiting the scope of protection of this invention.

[0056] Figure 1This is the overall flowchart of this method;

[0057] A rainfall prediction method based on marine radar imagery includes the following steps:

[0058] S1: Obtain the original radar data file, extract the azimuth of the obtained original radar data file, perform merging, interpolation and coordinate transformation processing, and call the display function to obtain the radar echo image; the radar echo image includes radar rain image and clear sky radar image.

[0059] S2: Differentiate the radar image during rainfall with the radar image during clear weather to remove central clutter interference;

[0060] S3: The radar rainfall image is reduced to a fixed resolution, and the image is segmented using the level set method; the fixed resolution can be 710*710 resolution;

[0061] S4: Calculate the optical flow field from N consecutive segmented radar echo images using the optical flow method;

[0062] S5: Combines radar echo images with motion vector fields to predict rainfall.

[0063] Steps S1, S2, S3, S4, and S5 are executed sequentially.

[0064] Further, the process of acquiring the raw radar data file involves extracting the azimuth of the acquired raw radar data file, performing merging and interpolation processing and coordinate transformation processing, and calling a display function to obtain a radar echo image; the radar echo image includes radar images during rainfall and radar images during clear weather; and includes the following steps:

[0065] S21: In rainy conditions, use marine radar to collect rainfall data via a data acquisition card and obtain the original radar data file;

[0066] S22: The acquired radar raw data file is subjected to azimuth extraction. The highest two bits of the sampled data are the azimuth reference and the azimuth data, respectively. The transition between 0 and 1 in the highest bit indicates that the antenna has turned past the azimuth reference. Then, the azimuth data information of the second highest bit is extracted to realize the formatting of the acquired data information.

[0067] S23: The formatted data information is merged and interpolated. Under different ranges, the correspondence between azimuth pulses and trigger pulses may deviate. If one azimuth pulse corresponds to multiple trigger pulses, the first echo information is retained. If one azimuth pulse does not have a corresponding trigger pulse, the previous echo information is copied to this azimuth pulse to ensure that each azimuth has corresponding echo information.

[0068] S24: Perform coordinate transformation on the merged and interpolated data. The radar information is in polar coordinates, so convert the polar coordinates to a rectangular coordinate system, then generate matrix data and store the matrix data file.

[0069] S25: Then read the matrix data file, save it to the corresponding matrix location in memory, convert the azimuth distance matrix to Cartesian coordinates, store it in a new matrix, and then call the display function to display the radar echo image.

[0070] Further: The step of differentiating the radar rainfall image and the clear sky radar image in the radar echo image includes the following steps:

[0071] S31: Because the buildings and school buildings near the radar installation location are relatively fixed, such clutter has a strong reflective effect on radar electromagnetic waves, forming high-intensity reflected echoes. Therefore, under clear weather conditions, radar data with the same acquisition range and gain as the radar is collected and converted into radar echo images.

[0072] S32: By subtracting the corresponding pixel values ​​of the obtained radar rainfall image from those of the clear sky radar image, clutter interference is removed, leaving only the rain area information for subsequent processing.

[0073] Further: The step of reducing the radar rainfall image to a fixed resolution and segmenting the radar echo image using the level set method includes the following steps:

[0074] S41: Reduce the radar rainfall image to 710*710 resolution for easier operation below;

[0075] S42: Set initial parameters λ1, λ2, α1 weighting constant and α2, maximum number of iterations IterNum, edge detection matrix g.

[0076] S43: Initialize the pseudo-level set, set one part to u(x) > 0.5, and the other part to u(x) < 0.5. u(x) = 0.5 is the contour, which does not need to be operated.

[0077] S44: Design two fitted images, a globally blurred fitted image and a locally blurred fitted image, and construct an energy function that includes the global term and the fitted term.

[0078] S45: Calculate the energy of the entire image domain, update the parameters of the entire image domain, calculate the new energy of the entire image domain, change the membership degree, and use a Gaussian filter to regularize and smooth the pseudo-level set function.

[0079] S46: Repeat step S45 until the iteration is completed after IterNum iterations.

[0080] S47: After image segmentation, add a module to retain image data and preserve the information of rainfall areas.

[0081] Furthermore, the calculation of the motion vector field from N consecutive segmented radar echo images using the optical flow method includes the following steps:

[0082] The optical flow field of N radar echo images is calculated using the LK optical flow method, where N≥10.

[0083] Track the sequence of N images. Assuming that the optical flow is a constant in the neighborhood of a pixel, use the least squares method to solve the basic optical flow equation for all pixels in the neighborhood to generate a smooth motion vector field. The basic optical flow equation is shown in formula (1).

[0084]

[0085] in, and The instantaneous velocity of that point in the x and y directions, also known as optical flow, This represents the change of a pixel over time.

[0086] Furthermore, the method of combining radar echo images with motion vector fields to predict rainfall includes the following steps:

[0087] By combining the deterministic components and perturbation terms of Lagrange persistence, precipitation is dynamically scaled.

[0088] An autoregressive (AR) process is used to calculate the evolution of the precipitation field, and the radar echoes are extrapolated using an advection scheme to predict rainfall.

[0089] A rainfall prediction method based on marine radar imagery includes:

[0090] Acquisition module: Used to acquire raw radar data files;

[0091] The acquisition module is used to extract the azimuth of the acquired raw radar data file, perform merging, interpolation, and coordinate transformation, and call the display function to obtain the radar echo image; the radar echo image includes radar images during rainfall and radar images during clear weather.

[0092] Differential module: used to differentiate the radar rainfall image from the radar echo image with the clear sky radar image to remove central clutter interference;

[0093] Segmentation module: Used to reduce radar rainfall images to a fixed resolution and segment radar echo images using the level set method;

[0094] Calculation module: used to calculate the optical flow field from N consecutive segmented radar echo images using the optical flow method;

[0095] Prediction module: Used to combine radar echo images with motion vector fields to predict rainfall.

[0096] Example 1

[0097] First, rainfall images are acquired, resulting in the image shown below. Then, the rainfall images are subjected to image difference between sunny and rainy days. Next, the difference image is segmented. After segmentation, the motion vector field is calculated from 10 consecutive radar echo images. Finally, rainfall prediction is performed.

[0098] Figure 2 This is a radar echo image of the rainfall.

[0099] Figure 3 Differential radar echo plots for sunny and rainy days.

[0100] Figure 4 This is a segmentation diagram of the radar echo image.

[0101] Figure 5 This is an extrapolation diagram of the motion vector field.

[0102] Figure 6 This is a rainfall forecast map for 30 minutes later.

[0103] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A rainfall prediction method based on marine radar imagery, characterized in that: Includes the following steps: The radar raw data file is acquired, and its azimuth is extracted. Then, it is merged, interpolated, and transformed. Finally, a display function is called to obtain the radar echo image. The radar echo image includes radar images during rainfall and radar images during clear weather. The radar image during rainfall in the radar echo image is compared with the radar image during clear weather to remove central clutter interference. The radar precipitation image is reduced to a fixed resolution, and the radar echo image is segmented using the level set method. The motion vector field is calculated from N consecutive segmented radar echo images using the optical flow method. Rainfall can be predicted by combining radar echo images with motion vector fields; The process of downscaling radar rainfall images to a fixed resolution and segmenting radar echo images using the level set method includes the following steps: S41: Reduce radar rainfall images to a fixed resolution; S42: Set initial parameters , , Weighted constants and Maximum number of iterations Edge detection matrix ; S43: Initialize the pseudo-level set, setting a portion of the pseudo-level set to... The other pseudo-level sets are set to ; S44: Design two fitting images: a globally blurred fitting image and a locally blurred fitting image, and construct an energy function that includes global and fitting terms; S45: Calculate the energy of the entire image domain, update the parameters of the entire image domain, calculate the new energy of the entire image domain, change the membership degree, and use a Gaussian filter to regularize and smooth the pseudo-level set function. S46: Repeat step S45 until the iteration is completed after IterNum iterations; S47: After image segmentation, information about the rainfall area is preserved; The method of combining radar echo images with motion vector fields to predict rainfall includes the following steps: By combining the deterministic components and perturbation terms of Lagrange persistence, precipitation is dynamically scaled. An autoregressive (AR) process is used to calculate the evolution of the precipitation field, thereby enabling rainfall prediction.

2. The rainfall prediction method based on marine radar imagery according to claim 1, characterized in that, The process of acquiring the raw radar data file involves extracting its azimuth, performing merging and interpolation, and coordinate transformation. A display function is then called to obtain the radar echo image. The radar echo image includes both rain-related and clear-sky radar images. The process includes the following steps: S21: In rainy conditions, use marine radar to collect rainfall data via a data acquisition card and obtain the original radar data file; S22: The acquired radar raw data file is subjected to azimuth extraction. The highest two bits of the sampled data are the azimuth reference and the azimuth data, respectively. The transition between 0 and 1 in the highest bit indicates that the antenna has turned past the azimuth reference. Then, the azimuth data information of the second highest bit is extracted to realize the formatting of the acquired data information. S23: The formatted data information is merged and interpolated. Under different ranges, the correspondence between azimuth pulses and trigger pulses may deviate. If one azimuth pulse corresponds to multiple trigger pulses, the first echo information is retained. If one azimuth pulse does not have a corresponding trigger pulse, the previous echo information is copied to this azimuth pulse to ensure that each azimuth has corresponding echo information. S24: Perform coordinate transformation on the merged and interpolated data. The radar information is in polar coordinates, so convert the polar coordinates to a rectangular coordinate system, then generate matrix data and store the matrix data file. S25: Then read the matrix data file, save it to the corresponding matrix location in memory, convert the azimuth distance matrix to Cartesian coordinates, store it in a new matrix, and then call the display function to display the radar echo image.

3. The rainfall prediction method based on marine radar imagery according to claim 1, characterized in that, The step of differentiating the radar rainfall image from the radar echo image with the clear sky radar image includes the following steps: S31: Under clear weather conditions, collect radar data with the same range and gain as the radar, and convert it into a radar echo image; S32: By subtracting the corresponding pixel values ​​of the radar rainfall image from those of the clear sky radar image, clutter interference is removed, leaving only the information of the rain area.

4. The rainfall prediction method based on marine radar imagery according to claim 1, characterized in that, The calculation of the motion vector field from N consecutive segmented radar echo images using the optical flow method includes the following steps: S51: Calculate the optical flow field from N radar echo images using the LK optical flow method; S52: Track a sequence of N images to generate a smooth motion vector field.

5. A rainfall prediction method based on marine radar imagery according to claim 1, characterized in that, The N≥10.

6. A rainfall prediction device based on marine radar images, characterized in that: include: Acquisition module: Used to acquire raw radar data files; The acquisition module is used to extract the azimuth of the acquired raw radar data file, perform merging, interpolation, and coordinate transformation, and call the display function to obtain the radar echo image; the radar echo image includes radar images during rainfall and radar images during clear weather. Differential module: used to differentiate the radar rainfall image from the radar echo image with the clear sky radar image to remove central clutter interference; Segmentation module: Used to reduce radar rainfall images to a fixed resolution and segment radar echo images using the level set method; The process of downscaling radar rainfall images to a fixed resolution and segmenting radar echo images using the level set method includes the following steps: S41: Reduce radar rainfall images to a fixed resolution; S42: Set initial parameters , , Weighted constants and Maximum number of iterations Edge detection matrix ; S43: Initialize the pseudo-level set, setting a portion of the pseudo-level set to... The other pseudo-level sets are set to ; S44: Design two fitting images: a globally blurred fitting image and a locally blurred fitting image, and construct an energy function that includes global and fitting terms; S45: Calculate the energy of the entire image domain, update the parameters of the entire image domain, calculate the new energy of the entire image domain, change the membership degree, and use a Gaussian filter to regularize and smooth the pseudo-level set function. S46: Repeat step S45 until the iteration is completed after IterNum iterations; S47: After image segmentation, information about the rainfall area is preserved; Calculation module: used to calculate the optical flow field from N consecutive segmented radar echo images using the optical flow method; Prediction module: Used to combine radar echo images with motion vector fields to predict rainfall; The method of combining radar echo images with motion vector fields to predict rainfall includes the following steps: By combining the deterministic components and perturbation terms of Lagrange persistence, precipitation is dynamically scaled. An autoregressive (AR) process is used to calculate the evolution of the precipitation field, thereby enabling rainfall prediction.

Images